Introduction

Many changes in thyroid hormones occur as a result of illness or nutritional deprivation. These changes consist of decreased levels of serum triiodothyronine (T3) and/or thyroxine (T4), without an increase in the thyroid-stimulating hormone (TSH) level [1, 2]. The combination of these findings is termed non-thyroidal illness syndrome (NTIS), euthyroid sick syndrome, or low T3 syndrome, indicating a systemic disease outside the thyroid that causes abnormal levels of thyroid hormones and is often considered a compensatory mechanism for the body. NTIS has been reported in patients with acute and chronic illnesses, including infectious diseases, cardiovascular and gastrointestinal diseases, cancer, and trauma [3, 4], which are quite common in patients in intensive care units (ICUs) [5, 6].

Older patients often have multiple chronic diseases and a poor nutritional status; thus, NTIS is quite common in these patients. Tognini [7] reported that among older patients (≥ 65 years of age) hospitalised for acute illness, the prevalence of NTIS was 31.9% and significantly associated with acute renal failure, New York Heart Association classification IV heart failure, and metastatic cancer. In previous studies, NTIS showed a high sensitivity and specificity for predicting patient mortality in ICU patients [8]. Serum T3 levels further decrease as the severity of disease progresses. However, among hospitalised older patients on general wards, data on the association of NTIS and mortality are lacking.

In this study, we analyse clinical data from inpatients in advanced age with or without NTIS, and their long-term outcomes for 7 years, to investigate the association of NTIS and mortality.

Patients and methods

Study design

This study collected data from patients admitted to the First Medical Center and the Second Medical Center, Chinese PLA General Hospital retrospectively, and recorded their outcome in 7 years follow-up. The patients in our hospital with long term follow-up were military veterans, therefore, there were few female patients, the study only included male patients to reduce bias. The inclusion criteria for the study were as follows: (1) aged 60 years or above, (2) hospitalised between January 2011 and December 2012, because the electronic medical record in our hospital can be obtained from 2011, (3) admitted on general internal medicine wards and (4) complete data on thyroid function obtained within 1 day of admission. The exclusion criteria were as follows: (1) diagnosed with thyroid disease such as hyperthyroidism, hypothyroidism, subclinical hyperthyroidism, subclinical hypothyroidism, or Hashimoto's thyroiditis, (2) history of thyroid surgery, (3) taking amiodarone, levothyroxine, glucocorticoid, dopamine, or interferon. A total of 8552 patients were admitted to the hospital from January 2011 to December 2012, and 959 of whom were enrolled in the present study and followed up for 7 years, until December 2019 (see Fig. 1). A total of 35 patients (3.7%) were lost to follow up including 8 patients in the NTIS group, and 27 patients in the non-NTIS group for migration or the change of medical designated hospitals.

Fig. 1
figure 1

Flowchart illustrating the inclusion/exclusion of individuals in this study

Full size image

Venous blood samples were collected early in the morning after an overnight fast on the second day after admission, including thyroid hormone (total T4, triiodothyronine, free T3, free T4, TSH), biochemical indicators (total protein (TP), albumin, prealbumin (PA), urea nitrogen (UN), fasting blood glucose (FBG), blood lipid, creatinine (Cr), uric acid (UA), alanine transaminase (ALT), aspartate transaminase (AST), and haemoglobin (Hb)). Radioimmunoassay was used to measure the serum levels of triiodothyronine, free T3, total T4, free T4, and TSH. The Sysmex Xt-1800 Automated Hematology Analyzer (SYSMEX Corporation, Japan) was used for routine blood tests. Biochemistry measurements were performed using i-CHROMA Reader (Boditech Med Inc. Korea). Primary nursing care is for (1) critically or severely ill patients with unstable vital signs; (2) patients at high risk of cardiovascular events or other life-threatening events; (3) patients cannot take care of themselves. Secondary/tertiary nursing care are for patients with stable vital signs and can partially or completely take care of themselves. The time and cause of death were reported for all deceased patients by doctors.

The study was conducted in accordance with the ethical rules of the Helsinki Declaration. The study protocol was approved by the Ethics Committee of Chinese PLA General Hospital.

Indicators and outcomes

The subjects were divided based on the presence of NTIS into the NTIS group and normal thyroid function (non-NTIS) group. Decreased serum triiodothyronine and/or free T3, normal or mildly reduced total T4 or free T4, and normal TSH levels were the diagnostic criteria for NTIS [9]. Normal thyroid hormone levels are as follows: total T4 (55.34–160.88 nmol/L), triiodothyronine (1.01–2.95 nmol/L), free T3 (2.76–6.30 pmol/L), free T4 (10.42–24.32 pmol/L), and TSH (0.35–5.50 uIU/mL). Intra-batch and batch-to-batch variations were < 3.35% and < 5.04%, respectively. The outcome was defined as all-cause mortality during the 7 years follow-up.

Statistical analysis

Normally distributed data are expressed as means ± standard deviation, non-normally distributed data as medians and quartiles, and category data are express as frequency and percentage. Normally distributed data were compared between the two groups using t test or t' test. The rank sum test was used to compare the data distribution between the two groups. Chi square test was used to compare the difference of category data. Overall survival at 7-year was estimated using the Kaplan–Meier method, and log-rank test was used to test the difference of mortality between two groups. To further evaluate the associations of thyroid hormone levels with all-cause mortality, bivariate and multivariate Cox proportional hazards models were used. Hazard ratio (HR) and 95% confident interval (95% CI) of each model was reported.

SPSS 24.0 and R was used for the statistical analysis. All data were evaluated using two-sided tests. Statistical significance was set at P < 0.05.

Results

Comparison of baseline data between the NTIS and non-NTIS groups

Most patients in both groups were in advanced age, and patients in the NTIS group were older (87.9 ± 6.0 vs. 85.8 ± 8.5, P < 0.01). Compared to the non-NTIS group, triiodothyronine, free T3, total T4, free T4 and TSH levels were lower, moreover, respiratory disease was the more common cause of hospitalization and the length of hospitalization was longer in the NTIS group. The percentage of patients accompanied by respiratory disease and chronic kidney disease were more in the NTIS group than in the non-NTIS group (P < 0.01). TP, albumin, PA, Hb, UA and high-density lipoprotein cholesterol (HDL-C) levels were significantly lower (P < 0.05), while FBG and UN levels were higher, in the NTIS group than in the non-NTIS group (P < 0.01), see Table 1.

Table 1 Comparison of baseline characteristics between the NTIS and non-NTIS groups
Full size table

Comparison of respiratory, cardiovascular, kidney and tumor disease severity between the NTIS and non-NTIS groups

The percentage of patients in primary nursing care was more in the NTIS group than in the non-NTIS group (P < 0.01). Furthermore, compared with the non-NTIS group, there were more patients accompanied by respiratory failure and renal dysfunction on haemodialysis in the NTIS group (P < 0.01), but no significant differences were observed in patients with heart failure or metastatic cancer (Table 2).

Table 2 Comparison of disease severity between the NTIS and non-NTIS groups
Full size table

Comparison of 1-year, 3-year and 7-year mortality between the NTIS and non-NTIS groups

During the 1-year follow-up period, 131 patients died: 55 (27.5%) in the NTIS group and 76 (10.0%) in the non-NTIS group (P < 0.01). At 3-years, the mortality was 46.0% in the NTIS group and 21.3% in the non-NTIS group (P < 0.01). A total of 622 patients died during the 7-year follow-up period, with 152 in the NTIS group and 470 in the non-NTIS group (P < 0.01, Fig. 2a). At 7-years, the survival rate was 24.0% in the NTIS group and 38.1% in the non-NTIS group (Fig. 2b). Furthermore, Kaplan–Meier survival analysis according to tertiles, showed that 7-year survival rate in the NTIS group obviously reduced with the decrease of free T3 (P < 0.01, Fig. 3a), and 7-year survival rate reduced with the increase of free T4 (P < 0.01, Fig. 3b).

Fig. 2
figure 2

Comparison of all-cause mortality and 7-year Kaplan–Meier survival curve of the NTIS and non-NTIS groups. a Comparison of mortality at 1 year, 3 years and 7 years in the NTIS and non-NTIS groups. b Seven-year Kaplan–Meier survival curve of the NTIS and non-NTIS groups

Full size image
Fig. 3
figure 3

Seven-year Kaplan Meier survival curves according to tertiles of free T3 (a) and free T4 (b) in the NTIS group

Full size image

Associations of thyroid hormone levels with all-cause mortality according to Cox proportional hazards models

Cox proportional hazards models showed that the reduced triiodothyronine and free T3 levels were significantly associated with higher mortality (P < 0.01). However, after adjustment for albumin and Hb, triiodothyronine and free T3 levels were not associated with all-cause mortality. Furthermore, lower free T4 level was shown to be associated with reduced all-cause mortality with a HR (95% CI) of 0.91 (0.88, 0.94), even after adjusting for all other confounding factors (P < 0.01; Table 3).

Table 3 Cox proportional hazards models of the associations of thyroid hormone levels with all-cause mortality
Full size table

Discussion

In the older population, thyroid hormone levels can help monitor health status, predict short-term and long-term clinical prognoses, predict disease severity, and assess quality of life and survival status. In previous studies, the frequency of thyroid dysfunction increased with advancing age in the hospitalised older patients. The prevalence of NTIS in hospitalised severely or debilitated older patients can be as high as 32–62% [8, 10, 11]. In the present study, we investigated 959 male patients in advanced age hospitalised for various reasons with a long follow-up for 7 years. We found that there were 200 patients with NTIS (20.9%) whose albumin levels were lower, and the 7-year survival rate was significantly lower in the NTIS group than in the non-NTIS group; in addition, a reduced free T3 level with low albumin and Hb levels was associated with all-cause mortality, and a higher free T4 in the normal range was also a strong predictor for long-term mortality risk in hospitalised older patients.

NTIS is often associated with nutritional deficiencies or acute and chronic diseases. Protein and UA levels are indicators of nutritional status. Proteins also play an important role in the synthesis and transport of thyroid hormones. Previous studies showed that in patients with NITS, the serum albumin level was reduced, and the free T3 level was positively correlated with the albumin level [12, 13]. In this study, the albumin level was also significantly reduced in the NTIS group compared with the non-NTIS group. It is supposed that decreased albumin level leads to a decrease in the conversion of T4 to T3, resulting in a decrease in T3 levels or a decrease in T4 binding to the protein, which accelerates the removal of thyroid hormones [14]. In the present study, a reduced free T3 level with low albumin and Hb levels was associated with all-cause mortality which further confirming that malnutrition is closely related with NTIS.

NTIS is a common thyroid dysfunction in CKD patients [12, 15], and its mechanism is associated with the kidney's involvement in the synthesis, secretion, and metabolism of thyroid hormones. In kidney disease, chronic metabolic acidosis and inflammatory factors lead to the inhibition of deiodinase activity, and the conversion of T4 to T3 in kidney tissue and other tissue types is reduced [16]. Hypothalamic-pituitary-thyroid axis dysfunction combined with loss of T4 in the urine causes triiodothyronine and total T4 levels decreased [17]. A decrease in the glomerular filtration rate (GFR) reduces iodine excretion, resulting in an iodine-blocking effect (Wolff-Chaikoff effect) [17]. Song et al. [18] retrospectively analysed 2,284 subjects with normal TSH levels and found that as the estimated GFR (eGFR) decreased in CKD patients, the prevalence of low T3 syndrome gradually increased; the eGFR was positively correlated with the serum T3 level independent of age and serum protein levels. In patients with chronic haemodialysis, reduced free T3 levels were a strong predictor of all-cause mortality [19]. Similar with previous studies, we found that among hospitalised older patients, the UN level in the NTIS group were higher, and there were more patients accompanied by chronic kidney disease, especially renal dysfunction on haemodialysis in the NTIS group than in the non-NTIS group.

NTIS also has been reported in patients with chronic obstructive pulmonary disease (COPD) accompanied by severe hypoxemia, and patients with obstructive sleep apnea accompanied by severe nocturnal hypoxemia [20, 21]. Moreover, recent research found that NTIS was common in mild-to-moderate COVID-19 patients and could predict clinical deterioration independent of SARS-CoV-2 viral load, age, inflammatory indicators and tissue injury [22]. Hypoxemia causes hypothalamic-pituitary-thyroid axis impairment and hypercapnia disrupts free T3 and free T4 production, which may be the main mechanism in NTIS development [20]. Our data demonstrated that more hospitalised older patients suffered from respiratory disease and even respiratory failure in the NTIS group than in the non-NTIS group.

In this study, Kaplan–Meier survival analyses showed that the survival rate was significantly lower in the NTIS group than in the non-NTIS group. Cox proportional hazards models showed that the reduced free T3 level increased all-cause mortality, however, after adjustment for albumin and Hb, free T3 levels did not predict mortality which indicated that patients with nutritional deficiencies presenting with lower free T3 level had poor prognosis. Unexpectedly, lower free T4 level was shown to be associated with reduced all-cause mortality even after the adjustment for all other confounding factors which suggested that a higher free T4 in the normal range might be a predictor for mortality risk in hospitalised older patients. At present, the association between thyroid hormone levels and mortality in older people are inconsistent. Some studies demonstrated that low free T3 levels were associated with all-cause mortality in patients with acute heart failure or admitted to ICUs [23, 24]. However, some other studies found that free T3 levels were non-specifically related with mortality after adjustment for potential confounders [25], whereas free T4 positively associated with death in advanced age [26]. Rozing et al. found that offspring of nonagenarian siblings had lower serum free T4 levels compared with their partners and reduced free T4 levels are associated with familial longevity [27]. In a study of 3885 community-dwelling men aged 70–89 years with a follow-up of 6.4 ± 1.5 years, Yeap et al. reported that higher free T4 levels are associated with increased all-cause mortality in euthyroid older men, after accounting for age, smoking, physical factors, and medical comorbidities [28]. Another retrospective cohort analysis with a mean follow-up of 3 years also showed that free T4 level is an independent predictor for mortality risk in hospitalized NTIS patients with chronic diseases [29]. Similar with above studies, a higher free T4 level in the normal range was associated with a worse prognosis among hospitalised older patients with long-term follow-up in the present study. Moreover, Kaplan–Meier survival analysis according to tertiles showed that survival rate at 7-year reduced with the increase of free T4. However, the underlying mechanism is unclear, which may be the consequence of low activity of deiodinase that converts T4 into the biologically active T3, and the details need to be further investigated.

The present study had several limitations including failure to evaluate many factors such as functional, nutritional, cognitive, frailty information that affect the patient prognosis. In this retrospective cohort study, the treatment protocols and response to the treatment may also have affected the patients’ outcomes. Due to the small sample size, many influencing factors were difficult to quantify, and the patients were not stratified according to the above-mentioned factors. In addition, only older male inpatients were analysed. Whether the study results can be generalised to the general population need to be further investigated.

In conclusion, the 7-year survival rate was lower in the NTIS group than non-NTIS group among older male patients on general wards. Patients with nutritional deficiencies presenting with free T3 level had poor prognosis, and a higher free T4 in the normal range was also a strong predictor for long-term mortality risk in hospitalised older male patients.